Emotional aspects play an important part in our interaction with music. However, modelling these aspects in MIR systems have been notoriously challenging since emotion is an inherently abstract and subjective experience, thus making it difficult to quantify or predict in the first place, and to make sense of the predictions in the next. In an attempt to create a model that can give a musically meaningful and intuitive explanation for its predictions, we propose a VGG-style deep neural network that learns to predict emotional characteristics of a musical piece together with (and based on) human-interpretable, mid-level perceptual features. We compare this to predicting emotion directly with an identical network that does not take into account the mid-level features and observe that the loss in predictive performance of going through the mid-level features is surprisingly low, on average. The design of our network allows us to visualize the effects of perceptual features on individual emotion predictions, and we argue that the small loss in performance in going through the mid-level features is justified by the gain in explainability of the predictions.

There has been rapid progress recently on the application of deep networks to the solution of partial differential equations, collectively labelled as Physics Informed Neural Networks (PINNs). In this paper, we develop Physics Informed Extreme Learning Machine (PIELM), a rapid version of PINNs which can be applied to stationary and time dependent linear partial differential equations. We demonstrate that PIELM matches or exceeds the accuracy of PINNs on a range of problems. We also discuss the limitations of neural network based approaches, including our PIELM, in the solution of PDEs on large domains and suggest an extension, a distributed version of our algorithm -{}- DPIELM. We show that DPIELM produces excellent results comparable to conventional numerical techniques in the solution of time-dependent problems. Collectively, this work contributes towards making the use of neural networks in the solution of partial differential equations in complex domains as a competitive alternative to conventional discretization techniques.

In this paper we introduce a numerical method for parabolic PDEs that combines operator splitting with deep learning. It divides the PDE approximation problem into a sequence of separate learning problems. Since the computational graph for each of the subproblems is comparatively small, the approach can handle extremely high-dimensional PDEs. We test the method on different examples from physics, stochastic control, and mathematical finance. In all cases, it yields very good results in up to 10,000 dimensions with short run times.

Knowledge distillation (KD) is a technique to derive optimal performance from a small student network (SN) by distilling knowledge of a large teacher network (TN) and transferring the distilled knowledge to the small SN. Since a role of convolutional neural network (CNN) in KD is to embed a dataset so as to perform a given task well, it is very important to acquire knowledge that considers intra-data relations. Conventional KD methods have concentrated on distilling knowledge in data units. To our knowledge, any KD methods for distilling information in dataset units have not yet been proposed. Therefore, this paper proposes a novel method that enables distillation of dataset-based knowledge from the TN using an attention network. The knowledge of the embedding procedure of the TN is distilled to graph by multi-head attention (MHA), and multi-task learning is performed to give relational inductive bias to the SN. The MHA can provide clear information about the source dataset, which can greatly improves the performance of the SN. Experimental results show that the proposed method is 7.05% higher than the SN alone for CIFAR100, which is 2.46% higher than the state-of-the-art.

The scope of research in the domain of activation functions remains limited and centered around improving the ease of optimization or generalization quality of neural networks (NNs). However, to develop a deeper understanding of deep learning, it becomes important to look at the non linear component of NNs more carefully. In this paper, we aim to provide a generic form of activation function along with appropriate mathematical grounding so as to allow for insights into the working of NNs in future. We propose ``Self-Learnable Activation Functions'' (SLAF), which are learned during training and are capable of approximating most of the existing activation functions. SLAF is given as a weighted sum of pre-defined basis elements which can serve for a good approximation of the optimal activation function. The coefficients for these basis elements allow a search in the entire space of continuous functions (consisting of all the conventional activations). We propose various training routines which can be used to achieve performance with SLAF equipped neural networks (SLNNs). We prove that SLNNs can approximate any neural network with lipschitz continuous activations, to any arbitrary error highlighting their capacity and possible equivalence with standard NNs. Also, SLNNs can be completely represented as a collections of finite degree polynomial upto the very last layer obviating several hyper parameters like width and depth. Since the optimization of SLNNs is still a challenge, we show that using SLAF along with standard activations (like ReLU) can provide performance improvements with only a small increase in number of parameters.

Traditionally, the automatic recognition of human activities is performed with supervised learning algorithms on limited sets of specific activities. This work proposes to recognize recurrent activity patterns, called routines, instead of precisely defined activities. The modeling of routines is defined as a metric learning problem, and an architecture, called SS2S, based on sequence-to-sequence models is proposed to learn a distance between time series. This approach only relies on inertial data and is thus non intrusive and preserves privacy. Experimental results show that a clustering algorithm provided with the learned distance is able to recover daily routines.

We explore the problem of estimating the mass distribution of an articulated object by an interactive agent. Our method predicts the mass distribution accurately only using information that can be reliably acquired by the limited sensing and actuating capabilities of a robotic agent that interacts with it. Inspired by the role of exploratory play in human infants, we take the combined approach of supervised and reinforcement learning to train the agent such that it learns to strategically interact with the object for estimating its mass distribution. Our method consists of two neural networks: the policy network which decides how to interact with the object, and the predictor network that estimates the mass distribution given a history of observations and interactions. Using our method, we train a robotic arm to estimate the mass distribution of an object with moving parts (e.g. an articulated rigid body system) by pushing it on a surface with unknown friction properties. We also test the robustness of our learned model by transferring it to another robot arm with different end-effector geometry. The empirical results show that our method significantly outperforms the baseline agent which uses random pushes to collect data for estimation.

Multiple sequence to sequence models were used to establish an end-to-end multi-turns proactive dialogue generation agent, with the aid of data augmentation techniques and variant encoder-decoder structure designs. A rank-based ensemble approach was developed for boosting performance. Results indicate that our single model, in average, makes an obvious improvement in the terms of F1-score and BLEU over the baseline by 18.67% on the DuConv dataset. In particular, the ensemble methods further significantly outperform the baseline by 35.85%.

End-to-end automatic speech recognition (ASR) models are increasingly large and complex to achieve the best possible accuracy. In this paper, we build an AutoML system that uses reinforcement learning (RL) to optimize the per-layer compression ratios when applied to a state-of-the-art attention based end-to-end ASR model composed of several LSTM layers. We use singular value decomposition (SVD) low-rank matrix factorization as the compression method. For our RL-based AutoML system, we focus on practical considerations such as the choice of the reward/punishment functions, the formation of an effective search space, and the creation of a representative but small data set for quick evaluation between search steps. Finally, we present accuracy results on LibriSpeech of the model compressed by our AutoML system, and we compare it to manually-compressed models. Our results show that in the absence of retraining our RL-based search is an effective and practical method to compress a production-grade ASR system. When retraining is possible, we show that our AutoML system can select better highly-compressed seed models compared to manually hand-crafted rank selection, thus allowing for more compression than previously possible.

Learning algorithms become more powerful, often at the cost of increased complexity. In response, the demand for algorithms to be transparent is growing. In NLP tasks, attention distributions learned by attention-based deep learning models are used to gain insights in the models' behavior. To which extent is this perspective valid for all NLP tasks? We investigate whether distributions calculated by different attention heads in a transformer architecture can be used to improve transparency in the task of abstractive summarization. To this end, we present both a qualitative and quantitative analysis to investigate the behavior of the attention heads. We show that some attention heads indeed specialize towards syntactically and semantically distinct input. We propose an approach to evaluate to which extent the Transformer model relies on specifically learned attention distributions. We also discuss what this implies for using attention distributions as a means of transparency.